User interfaces for electronic devices traditionally engage a limited number of human senses. Users see images presented on displays, and hear sounds generated by speakers, but physical feedback has been limited to simple shakes or buzzes. Furthermore, in addition to having limited output capabilities, traditional mechanisms are bulky, rendering them difficult to incorporate into small form-factor devices.